Human genome integrity depends on many processes to ensure the fidelity of the duplication of DNA. The efficiency of these processes is crucial since errors in DNA can often be key to disease onset. An important process to insure genome integrity is the repair of damaged DNA. There are several types of DNA damage including (but not limited to): alkylation, oxidation, hydrolysis, adduct formation, base mismatch, among others. Alkylated DNA bases may be removed by two main routes: excision of the damaged base and activation of the base excision repair (BER) process, or direct dealkylation. The former route involves several enzymes involved in the BER cascade. The latter route may be performed by the AlkB family of enzymes. AlkB family enzymes are non-heme iron and ?-ketoglutarate dependent enzymes that perform an oxidative dealkylation of DNA. Some cancer treatments involve alkylating agents, and attempts have been made to enhance these therapies by inhibiting alkylating damage repair. Information gained from a detailed understanding of the structure and reaction mechanism of AlkB family proteins can aid in the development of inhibitors for these enzymes by providing useful information to develop transition state analogue inhibitors. One approach for this is via computational methods, including quantum mechanical/molecular mechanical (QM/MM) methods. Currently, most QM/MM implementations employ force fields that may not accurately describe the MM environment at close range, are not polarizable and lack methods to include long-range electrostatic effects. Our long-term goal is to understand the mechanism, structure and function of enzymes involved in DNA repair by means of computational simulations. To this end, the goals of the present proposal are: i) To study the structure/function/reactivity of AlkB family of enzymes by quantum mechanical/molecular mechanical (QM/MM), molecular dynamics (MD) and homology modeling. ii) To develop the first QM/MM program that interfaces a QM program with a two advanced force fields (GEM and AMOEBA) to accurately describe the MM environment; and to develop a novel method to introduce long-range electrostatic effects in QM/MM simulations. The detailed understanding of the structure, function and reaction mechanism of AlkB and its human homologues will provide insights into possible methods to inhibit these enzymes. Our collaborators, Prof. Robert Housinger and Prof. Thomas Hollis, will perform experimental studies based on our computational results. Profs. Pengyu Ren and David Case will provide assistance with the QM/MM implementation in the AMBER suite of programs. The successful completion of the proposed project will provide an accurate computational tool for the calculation of enzyme reactions, and the generation of structural and mechanistic insights on an important family of enzymes that may be used to enhance the efficacy of cancer treatments.